Exhaust Fabrication

[turboplanner]

11 minutes ago, facthunter said:

They are NOT tuned length. Also the way that engine is installed with little pipes everywhere a cracked exhaust pipe is not something you should encourage by a suss  repair . Weaken a section of pipe and that's where the crack will happen You don't even draw lines on them with a graphite pencil.   Nev

There's a simple tune on a piano, then someone comes along and adds slip notes grace notes, flowing chords and its a whole new sound.

If someone justs uses a simple formula to adjust all the pipes to the same optimum length, that's the simple tune. We beat them in racing.

[facthunter]

Tuned pipes only work well at certain revs. It's well known that the Rotax 9xx has mixture variations due to the  Inlet Manifolds pairing. It goes suck suck nothing nothing. . WE beat them in racing. Who's WE?

 

 You haven't been fooling with a fairground organ have you?  Nev

[turboplanner]

35 minutes ago, facthunter said:

Tuned pipes only work well at certain revs. It's well known that the Rotax 9xx has mixture variations due to the  Inlet Manifolds pairing. It goes suck suck nothing nothing. . WE beat them in racing. Who's WE?

 

 You haven't been fooling with a fairground organ have you?  Nev

You've given Skippy your advice; I've given him mine.

[T510]

It's not a Rotax designed exhaust, it is designed to fit the airframe the motor is mounted to. It's not a tuned length header.

 

With all the different airframes Rotax get mounted to I doubt many have the same header lengths as Skippy's application.

 

A quick google will show the same motor with dozens of different configurations including different muffler locations, twin mufflers, single outlets and twin outlets.

 

 

 

[skippydiesel]

No racing or engine "Hotting"  experience to speak of.

 

From my lifetime of engine maintenance and reading (others experince) I would guess that the typical (there is no actual standard) Rotax  912 exhaust installation, is quite some way from being  "optimised", even with all Rotax supplied component (as mine is).

 

Different length headers, log style 90 degree muffler, does not look to me, to be a tuned exhaust extraction installation. It does however, get the job done.

 

With the above in mind, my objective is to aim for reliability, which in this context includes having no part of the engine inadvertently contacting structural (engine frame) or non structural (cowling) part of the aircraft.

 

To this end I have installed "stiffer" engine mounts, which have themselves then been further stiffened, using Cupped Snubbing Washers and additional spacing washers (supplied by Sonex). These changes have almost but not quite, eliminated engine to frame contact, that likely occurs at engine shut down. Rotax 9's are infamous for having violent lateral movement, at shut down. This is usually attributed to the integral gear box. In my installation this may be exacerbated by having a relativly heavy 3 bladed, constant speed, propeller acting as a pendulum.

 

I feel that I only have two further cost effective solutions to try:

 

• Fabricate a new left side rear exhaust header (the offending component) with a wider obstruction (frame) avoiding curve - The most effective solution but quite costly in time & dollars.
• Create a shallow (10mm) dent in the existing header pipe, adjacent to the contact point. This with the aforementioned engine mount changes, should result in no further engine: frame contact. This is a simple (KISS), minimal down time/nil cost solution, that may have some minor impact on performance.

 

I will try the last solution first. To assess any changes in performance, I will compare future EGT readings with those already recorded.

 

[LoonyBob]

Physics is not black magic.

[turboplanner]

 

 

 

 

 

19 hours ago, skippydiesel said:

No racing or engine "Hotting"  experience to speak of.

None is needed - that was a red herring.

19 hours ago, skippydiesel said:

 

From my lifetime of engine maintenance and reading (others experince) I would guess that the typical (there is no actual standard) Rotax  912 exhaust installation, is quite some way from being  "optimised", even with all Rotax supplied component (as mine is).

This opens up two possibilities (a) The variants all meet the same standard as Rotax testing or (b) they don't. 

Since you don't want to reduce reliability from Rotax testing if (a) and you don't want to move even further away from the standard (b).

In either of those cases you don't want to deviate from the present pipe length.

Given that Rotax have supplied the lot you don't want to deviate from the present pipe length.

 

You can't apply 1940s theory and hope to match the reliability achieved by Rotax.

 

19 hours ago, skippydiesel said:

Different length headers, log style 90 degree muffler, does not look to me, to be a tuned exhaust extraction installation. It does however, get the job done.

You're not looking at a racing engine setup, nor do you need one.

 

[skippydiesel]

Turbs - You are a master of the out of context commentary - Simply put: most naturally aspirated Rotax 9 exhaust configurations, are unique to that aircraft or in factory supply , model. The main driver for exhaust configuration is, to fit within the available space (cowling/engine frame). Tuning (exhaust length) is a secondary consideration, that usually loses out to the former. There are a few exceptions - these tend to feature much convoluted custom supply (ie not Rotax) exhaust headers/muffler (expensive, heavy, complex) may have the muffler in atypical locations (eg above engine). Of the latter designs, the improvements in engine performance are likely to be relativly small for the cost involved (analogues to fitting a small aircraft with a CS prop)

[facthunter]

A lot of aircraft engines have poor design of intake and exhaust systems from the performance aspect which come a poor second to the constraints of limited space and weight.  and safety. You don't want a cracked exhaust anywhere under your cowl." Racing " valve timings (large overlap) require a more balanced intake and exhaust treatment. Fitting that type of power unit would require  exhaust plumbing  similar to what it had in the motorbike of skidoo or Jetski it came from to keep it's performance. It's VERY much more applicable to 2 stroke engines.. They depend onpressure waves and expansion chambers. Supercharged engines need it the least, where in the case of a turbocharger it as close to the heads as possible these days. and the charge is forced by a pump.     Nev

[LoonyBob]

14 hours ago, turboplanner said:

 

 

 

 

 

None is needed - that was a red herring.

This opens up two possibilities (a) The variants all meet the same standard as Rotax testing or (b) they don't. 

Since you don't want to reduce reliability from Rotax testing if (a) and you don't want to move even further away from the standard (b).

In either of those cases you don't want to deviate from the present pipe length.

Given that Rotax have supplied the lot you don't want to deviate from the present pipe length.

 

You can't apply 1940s theory and hope to match the reliability achieved by Rotax.

 

You're not looking at a racing engine setup, nor do you need one.

 

Turbs, you appear to miss the point: the PHYSICS of GAS FLOW, SHOCKWAVES, and INTERNAL COMBUSTION ENGINES has NOT changed since the 1930s. If you think it has, please point to the post-1930 "Law" of physics that did was not known prior to that.

 

 

[turboplanner]

6 hours ago, LoonyBob said:

Turbs, you appear to miss the point: the PHYSICS of GAS FLOW, SHOCKWAVES, and INTERNAL COMBUSTION ENGINES has NOT changed since the 1930s. If you think it has, please point to the post-1930 "Law" of physics that did was not known prior to that.

 

 

Correct, laws of physics don't change.

[spacesailor]

But the application of those laws have charged. 

the outboard Motor can be efficient on both slow speed as well as wot .revs . All to do with the exhaust tuned for different cycles.

spacesailor

[facthunter]

In a normal outboard there's no room for a  good tuned exhaust system. Quite a lot of them are 4stroke these days.  Nev

[skippydiesel]

Squashed the side of the offending header today. Could not perform operation in situ, so removed pipe. Used a piece of tight fitting water pipe, on the side I wanted to remain as is and Vice Grips to gradually squash in the point, at which it was contacting the engine frame. I did not need to use th socket as an anvil.

 

Ran the engine up to oil 50C. Shut it down as per usual. No contact mark on the witness tab. Great!

 

I  will fly the Sonex in a week or so, for a definitive test but very hopeful of this being the solution (along with all the erlier incremental efforts/improvements)

 

 

Edited February 10 by skippydiesel

[LoonyBob]

2 hours ago, spacesailor said:

But the application of those laws have charged. 

the outboard Motor can be efficient on both slow speed as well as wot .revs . All to do with the exhaust tuned for different cycles.

spacesailor

Heat engines are heat engines. Are we talking about uniflow two-stroke supercharged outboards, or disc-valved crankcase compression two strokes, or reed valve crankcase compression two strokes, or piston ported crankcase compression two strokes, or some form of four stroke?

All crankcase compression two-strokes* must open the exhaust port before the inlet/transfer port, and so they must close the exhaust port last. This allows some of the charge to escape into the exhaust. If said charge does not yet contain fuel (a la Sarich "Orbital Engine Combustion technology", which is a rehash of airblast injection from pre-WW1), the only penalty is a depressed scavange pressure. However, as the crankshaft bearings must be lubricated by an oil mist, some oil escapes.

If the charge contains fuel, the only practicable system is, as Scott used in 1908, to reflect the previous exhaust pressure wave** within the exhaust, such to reverse the terminal exhaust flow at certain (narrow) bands in the rev range. In such circumstances, the escaping charge (fuel, oil & air) is rammed back into the cylinder, avoiding wastage and achieving a slight supercharge. However, between the "ram" bands, the opposite effect takes place - the resonance draws extra charge out of the cylinder, and subcharges the cylinder.

 

*A very few early two-strokes used the crankcase compression to feed a manifold to timed valves at the combustion chamber, but the large clearance volume led to very low volumetric efficiencies.

**A travelling normal shockwave (a supersonic phenomena) (a) attenuates greatly at an open pipe end, which entry into a muffler simulates, and (b) constitutes a feeble ghost of a piston. Assuming a constant gas density and velocity adjacent to both sides of the open end, it is usual to assume a ~50% reflectivity. The first return to an open exhaust valve should be expected to have a noticeable pumping effect, but note that this first return would occur while the port was still discharging gas at a high rate (well before blow-through), and so would not automatically be of benefit. After 32 traverses of the zorst pipe (whilst waiting for blowthrough to start), assuming no losses in transit, the intensity would be reduced to 1/65,536th of its original value. Use of a resonance box, to reduce said attenuation, will help things, but not greatly at less than WOT...

The pressure wave, however, would be reflected very efficiently, though it would suffer edge attenuation by boundary layer formation, more than the shockwave does. However, after 32 traverses, boundary layer attenuation would be of the order of 15% loss.

Four strokes can also benefit, by using the odd numbered waves (which have negative pressure) to aid scavange. For a poppet valve engine (912/14/15/17/213?), the end of the valve guide and the stem would somewhat attenuated the closed end reflection, but even if the overall losses were ~50%, the WOT initial pressure is of the order of 4 bar...

 

The pressure wave effect is also present on the inlet side of the engine, although much less powerful; nevertheless, the "D" type Jag was able to lift its peak volumetric efficiency from ~87%(from memory) to over 120% by extending the inlet runner length (including the tract within the head) from ~8" to 48". This also reduced the volumetric efficiency between the power bands, but by tweaking the runner length, they were able to put the power bandes where the circuit and gearing needed them.

 

It is interesting to note that a conventional set of "headers" for a car engine, terminate in a 3-1 or 4-1 transition, before the muffler / "open end". This means that the shock and pressure waves are shared between 3 or 4 cylinders, not returned in toto to the cylinder of origin. The McLaren TAG engine used a device that destroys (consumes) shockwaves, and swallows pressure waves almost completely, yet produced ~850hp from 1.5 litres...

[spacesailor]

I was saying exhaust porting has changed , since the Old

' single ' tuned exhaust were being used on outboard motors .

spacesailor

 

[LoonyBob]

Okay, "powervalves" have become a thing; but their purpose is to allow engines with porting for huge power at high revs, to not stumble in the midrange. And yes, this does reduce the blowthrough (and emissions) under those conditions; but they don't change the fundamentals, unfortunately.

I'm a big fan of 2-strokes that do NOT use crankcase compression, but I'm going to have to build my own!

 

[facthunter]

 Hard to keep the Pistons cool when they fire each revolution..  I designed an opposing piston thing with variable compression but the standard things are going so well.  Nev

[skippydiesel]

 

Decided to "bite the bullet" and fix the problem, as I should have done, from almost the start of this saga. Way too many lazy/easy fix tried with only moderate improvement.

 

The aim; to increase the existing clearance (12 mm) between exhaust header and engine frame by about 10-12mm, giving a total clearance of 20-24mm

 

First stage - make a jig, using existing pipe, to hold the parts in the correct alignment.

 

I have decided to have a go at modifying what I already have. This will mean removing a small section of pipe just above the elbow coming out of the head. The remaining elbow can then be rotated. A slight bend will have to be created in the long straight section.  The resultant "gap" will need to be bridged/filled with a short length of elbow.

 

The hope is that the head & muffler ends will remain in the correct alignment.

 

Assistance/suggestions/comments - all welcome.

 

Edited March 10 by skippydiesel

[Thruster88]

Bolt some 3-5mm steel plates to your wooden jig and tack weld ends in three places before you cut. Tidying up the tacks at the end of job is easy and alignment is more assured.  

[skippydiesel]

Job done. Yet to be installed & tested. The bloke who did the work decided to remove most of the ceramic heat shield coating. Wondering what effect, if any, this might have on the EGT reading for this header (the other remains coated)

 

[skippydiesel]

Modified  (more clearance) header seems to be performing as intended  - did a little over an hour yesterday - all good!

[turboplanner]

5 minutes ago, skippydiesel said:

Modified  (more clearance) header seems to be performing as intended  - did a little over an hour yesterday - all good!

There you are, a design genius!

[spacesailor]

But why keep the round shape ( that you started with ).

Squashing it to an oval shape will lessen drag .

Many years ago , I made my own ' roof racks ' from 2"  stainless , that I picked up  heap . They looked shit when finished,  so I squashed that pipe to 1" by whatever it was . Then everyone wanted those great looking aerodynamic roof-racks .

I put a 3/4" wood dowel in the stainless pipe &used a press to get it right .

I have no idea what your curves would turn out after  pressing it .

spacesailor

 

[facthunter]

Once you could get Lucky, Lukey or Leaky.  Nev